Conventionally, in a closed-loop transmission diversity mode corresponding to OFDM (Orthogonal Frequency Division Multiplexing), when phase information or amplitude information is fed back for each subcarrier from a terminal to a base station, the following mapping is performed for each piece of information to reduce a necessary information amount.
A plurality of subcarriers is selected as reference subcarriers and remaining subcarriers are defined as adjacent subcarriers associated with the reference subcarriers. Phase estimate values and amplitude estimate values of the reference subcarriers are fed back based on normal mapping. Phase estimate values and amplitude estimate values of the adjacent subcarriers are fed back based on mapping with locations biased toward the vicinity of the phase estimate values and the amplitude estimate values of the associated reference subcarriers. On the other hand, in the base station, the phase estimate values and the amplitude estimate values fed back from the terminal are used to control a transmission antenna weight at the next transmission (see patent document 1).
On the condition that an information amount necessary for feedback is constant, performing such an operation may reduce the information amount in the closed-loop transmission diversity mode corresponding to OFDM by performing the above operations as compared to a mode of feeding back the phase information and amplitude information of all the subcarriers based on the normal mapping (the mapping applied to the reference subcarriers). As a result, the phases and amplitudes of the subcarriers may highly accurately be controlled and the performance of the transmission diversity may be improved without increasing the feedback information amount.
Patent Document 1: Japanese Patent Application Laid-Open No. 2003-87070, FIG. 4